In the current technology catalyst particles are immersed in oil that contains small amounts of dissolved hydrogen which results in catalyst surface being hydrogen starved and thus promotes the isomerisation to TFAs. The isomerization to TFA can be reduced by having a high hydrogen concentration at the surface of the catalyst. Our membrane reactor approach relies on adding hydrogen directly near the catalytic sites of known hydrogenation catalysts located on the skin of a polymeric high-performance asymmetric gas separation membrane. A membrane capable of selectively transporting hydrogen while acting to prevent any loss of liquid phase is incorporated in the reactor housing. Oil is pumped on one surface of the membrane where it comes into contact with the catalytic metal surface supported on the polymeric membrane support. The metal catalyst has a high hydrogen coverage that has diffused through the membrane due to an imposed chemical potential driving force. High concentrations of hydrogen on the catalyst surface, and the resulting decrease in temperature, promote the hydrogenation reaction at the expense of the cis to trans isomerization. The approach studied is also compatible with existing hydrogenation technologies and requires no additional utilities.
Experiments were performed with membranes having different membrane properties and catalyst loadings using platinum, palladium, and Pd-Pb alloy as catalyst. At an iodine value of 100 the membrane reactor (70C, 3.4 atm H2) produced less than 2 wt% TFA while the conventional system produced nearly 10 wt% TFA (Pt catalyst). The presentation will discuss the concept and the influence of membrane properties, catalyst type and catalyst loadings on TFA formation and hydrogenation selectivity.